Query-aware Hub Prototype Learning for Few-Shot 3D Point Cloud Semantic Segmentation

2025-12-09

AI Summary

Existing few-shot 3D point cloud semantic segmentation methods generate class prototypes solely from the support set, neglecting their semantic alignment with query data—leading to prototype bias and degraded generalization under distribution shift. To address this, we propose a query-aware hub prototype learning framework. First, we construct a cross-set bipartite graph between support and query sets to identify semantically critical support hub points, enabling query-relevant prototype generation. Second, we introduce a purity-weighted contrastive loss to refine prototype representations, mitigating boundary ambiguity and distribution misalignment. Our method achieves state-of-the-art performance on S3DIS and ScanNet, significantly outperforming prior approaches. It effectively narrows the semantic gap between prototypes and query samples, thereby enhancing generalization accuracy for novel categories in few-shot segmentation.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Semi-Supervised Learning
• Search and Optimization: Learning to Search

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Few-shot 3D point cloud semantic segmentation (FS-3DSeg) aims to segment novel classes with only a few labeled samples. However, existing metric-based prototype learning methods generate prototypes solely from the support set, without considering their relevance to query data. This often results in prototype bias, where prototypes overfit support-specific characteristics and fail to generalize to the query distribution, especially in the presence of distribution shifts, which leads to degraded segmentation performance. To address this issue, we propose a novel Query-aware Hub Prototype (QHP) learning method that explicitly models semantic correlations between support and query sets. Specifically, we propose a Hub Prototype Generation (HPG) module that constructs a bipartite graph connecting query and support points, identifies frequently linked support hubs, and generates query-relevant prototypes that better capture cross-set semantics. To further mitigate the influence of bad hubs and ambiguous prototypes near class boundaries, we introduce a Prototype Distribution Optimization (PDO) module, which employs a purity-reweighted contrastive loss to refine prototype representations by pulling bad hubs and outlier prototypes closer to their corresponding class centers. Extensive experiments on S3DIS and ScanNet demonstrate that QHP achieves substantial performance gains over state-of-the-art methods, effectively narrowing the semantic gap between prototypes and query sets in FS-3DSeg.

Problem

• Addresses prototype bias in few-shot 3D segmentation
• Models semantic correlations between support and query sets
• Refines prototypes to narrow semantic gap with queries

Innovation

• Query-aware Hub Prototype generation via bipartite graph
• Purity-reweighted contrastive loss for prototype refinement
• Explicit modeling of support-query semantic correlations